OK, I’m very new here. I jumped right in and submitted several “solutions” on 101 “The Finger”. I want to know what happened in the last hour of the vote. I want to be able to see how the runners up did.
It’s hard for me to see how this does not become a popularity contest of who has the most friends.
Apple:
Seeing previous votes: You’re absolutely right, there _should_ be a way of seeing what the votes were in the previous round. In fact I’ve made that feature #111 in the bug tracker. We’ll get on fixing that as soon as possible.
Popularity contest: This is a great point. One of the most valuable tasks the EteRNA community performs, perhaps the most valuable, is selecting to which RNAs we should devote our precious experimental resources. Any such voting mechanism can devolve into a “popularity contest,” which would very much be a shame. Provisionally, we hope to avoid this by asking the EteRNA community to respect the valuable experimental resources we’re putting at their disposal. Also, there’s inherent motivation to pick the right one, irrespective of your friends, because your lab score is proportional to how well your RNA synthesizes in practice.
However, we would also like to consider technical solutions to this problem, including modifications to the voting and scoring system.
We would welcome any thoughts or suggestions from you on this matter!
suggestions:
1 - no more than one vote per solution from any player (e.g., must spread out your 8 votes)
2 - no more than one solution per player - must pick your best (would keep the list to a manageable size)
Also, should you be able to vote for your own? I am kind of split on the idea. What does everyone else think?
Adding a feedback loop to voting such that, in addition to affecting lab score, the number of votes one gets in the future depends on results.
Apple: That’s a great idea.
We were thinking of changing the lab voting rules as follows.
-
Players can submit no more than one RNA.
-
Players can vote no more than once for any RNA.
-
Synthesized RNAs get scored according to the experimental results. (Call this the synthesis score.)
-
All RNAs get a design score. For synthesized RNAs, design score = synthesis score, but for non-synthesized RNAs, design score = synthesis score of the closest synthesized RNA (where distance is measured by the sequence distance).
-
Players receive the design score for any RNAs they design plus some fraction of that score if they voted for that RNA as well.
-
Optionally: The number of votes a player gets is proportional to his/her score.
Thoughts? Comments? Modifications?
We have also played around with some thoughts about whether we should distinguish between challenge and lab scores. Any thoughts on that would be appreciated as well.
Optionally: The number of votes a player gets is proportional to his/her score.
Not quite what I had in mind, unless the players score strongly reflects how well they have performed in recent past. If a player score can drop drastically after poor voting, then sure, but other wise I see to much voting power going to long term players.
Apple I really like your idea in terms of the feedback loop that effects how many votes you get.
The simplest version could be that you get +1 vote for the number of rounds in a row you have voted and designed. (for example: if you participated in the last 3 rounds, you get +3).
This would simply reward those people who are actively involved in the lab. And I don’t think it would favor long term players.
We could expand this dependent on how well your design (or designs you voted for did). So instead of +1 for each week, it’s +1*k where k is related to your synthesis scores in some way.
I think these two ideas would reward those players who consistently participate and actively participate by granting them more influence.
Is this along the lines of what you were thinking of Apple?
Alex. I like the idea of making it proportional to past performance, but having a time horizon, so that the entire game does not become dominated by some unbeatably high scoring players.